Research Article
Print
Research Article
An individual-based model of basking sharks in Ireland
expand article infoChelsea Gray‡§, Dale S. Rothman|, Erin Peters-Burton, Cynthia Smith‡, E.C.M Parsons#, Alexis Garretson¤«
‡ Department of Environmental Science and Policy, George Mason University, Fairfax, United States of America
§ Irish Basking Shark Group, Donegal, Ireland
| Department of Computational & Data Sciences, George Mason University, Fairfax, United States of America
¶ College of Education and Human Development, George Mason University, Fairfax, United States of America
# Centre for Ecology & Conservation, Penryn Campus, University of Exeter, Cornwall, United Kingdom
¤ The Jackson Laboratory for Mammalian Genetics, Bar Harbor, United States of America
« Graduate School of Biomedical Sciences, Tufts University, Boston, United States of America
Open Access

Abstract

Basking sharks (Cetorhinus maximus) are largely solitary, following highly individualized migration routes, yet they exhibit poorly understood aggregation behaviors, ranging from pairs to groups of up to 1,000 individuals. These surface aggregations put sharks at risk of boat strikes and are thought to be driven in part by localized zooplankton availability, though observational evidence suggests that food alone may not fully explain this behavior. Previous modeling approaches have successfully identified long-range basking shark migration trends but have not accurately reproduced behavior at finer spatial scales, and commonly used chlorophyll a proxies for zooplankton have produced inconsistent results when compared with sightings data.

This study presents the first individual-based model (IBM) of basking shark aggregation behavior, simulating shark movement around the Inishowen Peninsula, Ireland, from 1982 to 2018. Rather than relying on a phytoplankton proxy, zooplankton was simulated directly to capture fine-scale prey patchiness. Four behavioral submodels were tested and compared with three decades of sightings data from the Irish Basking Shark Group and Irish Whale and Dolphin Group: sharks seeking food only, sharks seeking other sharks only, sharks seeking both food and other sharks, and a random-movement control.

The combined food-and-social submodel produced the most realistic results, with lower error and greater qualitative correspondence to observed sightings data than the food-only, social-only, or random submodels, which rarely generated aggregations larger than two individuals. These findings indicate that basking shark aggregations emerge from a combination of environmental and social drivers rather than food availability or chance alone. This model demonstrates the viability of IBMs for simulating basking shark behavior, offers a flexible framework that can incorporate improved environmental and behavioral data as they become available, and provides a foundation for future research into the drivers of basking shark aggregation.

Key words:

Basking sharks, Individual-based model, marine conservation, shark behavior

Introduction

Basking sharks (Cetorhinus maximus) are the second-largest extant fish in the ocean, the only species in the family Cetorhinidae, and are considered endangered worldwide (Rigby et al. 2021). Basking sharks are often found feeding on zooplankton at the surface of coastal waters in the North Atlantic, where they are most at risk of anthropogenic threats, such as boat strikes and bycatch (OSPAR Commission 2015). Ireland is a notable hotspot of basking shark social, reproductive, and feeding activity, where they are sighted both alone and in groups at various locations along the coast (McInturf et al. 2026). Basking sharks are known to gather in large aggregations, ranging from two to several thousand individuals (Crowe et al. 2018; Sims et al. 2022). These aggregations are still poorly understood but regularly correlate with seasonal boat traffic, putting the sharks at risk of boat strikes (Speedie et al. 2009; Chapple et al. 2024). The purpose of these aggregations is still unknown, though proposed reasons include feeding, courtship, or both (Richardson et al. 2006; Speedie et al. 2009; Rudd et al. 2021a; Sims et al. 2022; Curtis et al. 2024). These aggregations include both feeding and non-feeding behaviors (torus swimming and parallel swimming) and are often in mixed-sex groups (Sims et al. 2022), though recent genetic research has also found that sharks may be more likely to surface near genetic relatives (Lieber et al. 2020).

In 2022, due partially to the significance of these aggregations, basking sharks were listed under Section 23 of Ireland’s Wildlife Act of 1976, forbidding disturbance of basking shark feeding and reproductive habits, such as these aggregations. Implementation of strong conservation policies requires an improved understanding of basking shark social behavior, which is critical for prioritizing areas of habitat for basking shark social, reproductive, or feeding potential. Currently, there is a lack of conclusive data surrounding basking shark social and migratory behavior, limiting the potential of temporal or area-based management measures. It is difficult to directly observe these aggregations or collect environmental data in situ, as doing so relies on finding an elusive species. Tagging data, while valuable, are prohibitively expensive and only provide data on a small subset of any population.

Here, the first individual-based model (IBM) of basking shark aggregation behavior is presented. The model’s capacity to reproduce observed aggregation patterns is demonstrated by simulating the period from 1982 to 2018 and comparing four behavioral submodels with three decades of sightings reports from the Irish Basking Shark Group (IBSG) and Irish Whale and Dolphin Group (IWDG).

While IBMs have been used for ecological research, they have never been applied to basking sharks (Gray et al. 2023). IBMs allow for bottom-up modeling, with populations that include individuals (i.e., simulated individual sharks) that are heterogeneous, whereas traditional models often assume uniformity or the “average” individual or environment (Uchmański and Grimm 1996). IBMs allow for the consideration of how individuals interact with and affect, or are affected by, their spatially explicit environment (Uchmański and Grimm 1996; Grimm and Railsback 2005). This allows for a stronger understanding of patterns of behavior, such as aggregations, and allows for increased environmental complexity and stochasticity (Squazzoni and Boero 2010). Understanding how food drives both long-term migratory patterns and localized basking shark behavior is vital for conservation in the face of climate change, and individual-based modeling is a mechanism for doing so. IBMs can be complementary to fieldwork, especially when fieldwork is prohibitively expensive and not always feasible due to unforeseen circumstances (e.g., COVID-19; Murphy et al. 2020), making it valuable to determine whether IBMs are suitable for simulating basking shark behavior.

Basking sharks are uniquely suited to an IBM because, although they are largely solitary and follow highly individualized migration routes, they exhibit poorly understood and understudied aggregation behaviors (Sims 2008; Rudd et al. 2021a; Sims et al. 2022). These surface aggregations, which appear seasonal, are thought to be driven in part by localized availability of zooplankton (Baduini 1995; Cotton et al. 2005; Okey et al. 2010; Crowe et al. 2018; Gore et al. 2018; Sims and Reid 2002a; Elcock et al. 2026). Previous methods of modeling have been successful at identifying long-range trends in basking shark migration but have yet to accurately reproduce basking shark behavior at a more localized scale (Doherty et al. 2017a; Braun et al. 2018; Austin et al. 2019; McInturf et al. 2022). Previous research utilizing chlorophyll a data, a commonly used proxy for zooplankton (Druon et al. 2019), has produced inconsistent trends when compared with basking shark sightings in Ireland and the United States (McInturf et al. 2022, 2026). This suggests that it is valuable to simulate zooplankton directly, especially on a small scale, rather than rely on phytoplankton as a proxy, to understand fine-scale basking shark patterns.

The IBSG and IWDG have collected publicly reported basking shark sightings for five decades (McInturf et al. 2026). These are reported by phone or internet form and include the latitude, longitude, date, time, and number of sharks spotted. In Ireland, basking sharks regularly come close to shore, so sightings reports are made by both those at sea (e.g., fishermen) and those on land (e.g., hikers). From 1982–2018, the IBSG/IWDG documented 2,136 sightings of sharks throughout Ireland. Nine hundred and sixty-nine (45%) of those included aggregations of two or more sharks (Irish Basking Shark Group, restricted access data, 1982–2024). One area of high basking shark activity is the Inishowen Peninsula (Co. Donegal; Fig. 1), where 458 sightings were reported to the IBSG/IWDG, 210 (45%) of which were of two or more sharks (Irish Basking Shark Group, restricted access data, 1982–2024). The Inishowen Peninsula includes an area (Malin Head & Lough Swilly) identified by the IUCN as an Important Shark and Ray Area for basking sharks (Jabado et al. 2025). It is also an area known for breaching basking sharks, a behavior hypothesized to be related to reproduction or courtship (Gore et al. 2018; Johnston et al. 2018; Rudd et al. 2021b). The Inishowen Peninsula sits within the Malin-Hebrides Sea, across from the Hebrides in Scotland, where tagging data indicate that the basking shark population in these two areas is shared, and basking sharks are documented to experience intermittent site fidelity in the region, as well as other parts of Ireland and Scotland (Doherty et al. 2017a; Johnston et al. 2019; Rudd et al. 2021a; Thorburn et al. 2024). The Hebrides were also recently designated as a marine protected area in part due to the basking shark aggregations, where both group feeding and courtship-like behaviors have been documented with camera tags (Speedie et al. 2009; Johnston et al. 2019; Marine Scotland 2020; Rudd et al. 2021a).

Figure 1. 

Model area. The Inishowen Peninsula, modeled in the IBM, is part of Ireland, adjacent to Northern Ireland. It sits in the Malin-Hebrides Sea, directly across from the Hebrides, Scotland.

The model presented here sought to simulate basking shark movement around Inishowen to determine whether IBMs can produce realistic behavioral results, in particular whether a more natural preyscape (simulated through zooplankton patchiness) can improve the realism and effectiveness of modeling shark behavior. The Inishowen Peninsula was selected due to its proximity to an established marine protected area and the availability of sightings data and to limit the simulated area for simplicity. The results of this model were compared with the IBSG/IWDG sightings dataset.

The IBM simulated daily basking shark movements around Inishowen under three behavioral conditions: 1) sharks only seek food, 2) sharks only seek other sharks, and 3) sharks seek both food and other sharks, to assess whether food availability, social attraction, or a combination of these were sufficient to reproduce the aggregations sighted around the Inishowen Peninsula. It is currently unknown where basking sharks mate or give birth, with only one recorded capture of a pregnant female in the North Atlantic in 1943 (Sims 2008). This IBM was an explanatory model, as described by Pilkey (2007). Rather than prediction, the model’s purpose was to understand a process or system, i.e., the drivers of the aggregations. For this research, an aggregation was defined as “the co-occurrence of two or more individuals in space and time due to the deliberate use of a common driver” (McInturf et al. 2023) and was measured by the presence of two or more sharks in a 1 km × 1 km patch. Testing these three behavioral hypotheses was conducted to demonstrate that the model behaves with a semblance of realism by testing the impact of key drivers (environmental, social) on shark aggregations alongside a random control. This research seeks to demonstrate the viability and value of utilizing IBMs as a method of simulating and understanding basking shark behavior.

Methods

The model was created in NetLogo (version 6.2.2, Wilensky 1999). Model development and testing were documented using the Transparent and Comprehensive Model Evaluation (TRACE; available upon request), Overview Design Document (ODD; Suppl. material 2: 1), and Objective Patterns Evaluation (OPE; Suppl. material 2: 2) methods (Grimm et al. 2006, 2014; Schmolke et al. 2010; Planque et al. 2022). Zooplankton data from the Continuous Plankton Recorder Survey (CPR), publicly available through the Archive for Marine Biological Association, were used to inform the zooplankton distribution in the model (Johns 2020). The model area’s visual map and sea depth were obtained from the General Bathymetric Chart of the Oceans (GEBCO Bathymetric Compilation Group 2020), a publicly accessible global bathymetry portal.

Basking shark sighting reports from the IBSG/IWDG were used for calibration and validation, as they constitute the only comprehensive dataset on basking shark population and distribution in Ireland (The Shark Trust 2025). The sightings database is a controlled-access database organized through a collaboration between the Irish Whale and Dolphin Group (IWDG) and Irish Basking Shark Group (IBSG), available upon request and subject to the data use policies of the IBSG/IWDG. The dataset also includes reports from the Marine Conservation Society (MCS) prior to 2007 (McInturf et al. 2026). The sightings data were originally captured on paper and are now collected online via a form that can be submitted by any member of the public (e.g., recreational and commercial boaters, observers on land) in both the Republic of Ireland and Northern Ireland. Each report includes the date, location (latitude/longitude), and number of animals observed. Each record from the IBSG/IWDG database was validated by the IWDG based on images of the animals, descriptions of their behaviors, and a priori knowledge of the local marine ecosystems (McInturf et al. 2026).

The primary objective of this research was to determine whether IBMs are a workable method for modeling basking shark behavior, demonstrating the viability of such a tool for future research and experimentation. This paper serves as a model description and proof-of-concept approach. To demonstrate the utility of the model, three hypotheses of behavior (sharks seek other sharks, sharks only seek food, and sharks seek a combination of food and other sharks) and a random control were tested using different submodels.

Model development and testing proceeded in four phases: (1) exploratory testing to establish plausible parameter ranges from the literature and small ad hoc trials; (2) parameter selection, in which candidate configurations were compared with sightings data using the mean error (ME), root mean square error (RMSE), and mean absolute error (MAE); (3) confirmatory model runs at full scale (200 sharks, 50 trials) across all four submodels; and (4) sensitivity analysis, robustness analysis, and a parametric stress test on the selected configuration. The IBM is available on GitHub (https://github.com/graychelsea13-byte/Basking_Shark_IBM/).

Model overview

Each time step in the model represents 24 h. The modeled area ranges from 56°N, -8°W to 55°N, -6.5°W, an area of 10,545 km2, with patches of 1 km × 1 km × 10 m (depth). Patches of 1 km × 1 km were selected to keep patch sizes small but manageable and to account for shark movements within 24 h (Doherty et al. 2017a). A depth of 10 m is the maximum estimated depth of the Continuous Plankton Recorder (Richardson et al. 2006). Though basking sharks are known to swim up and down within the water column (Rudd et al. 2021a; Elcock et al. 2026), simulated sharks remain in the top 10 m of water, as the IBSG/IWDG sightings reports are only of surfacing sharks. To simulate sharks either diving deeper than 10 m or leaving the model area, sharks disappear from the model (referred to as “migrate” in the model mechanism).

Each simulation depicts the period from 1 April–31 October1 for 1982–2018. This time of year was chosen because IBSG/IWDG sightings data show that this is when sharks are most likely to be surfacing in Ireland (McInturf et al. 2026). While some basking sharks may remain in Ireland during the winter, those that do likely remain well below a depth of 10 m (Doherty 2017) and, therefore, are unlikely to be accounted for in publicly reported sightings.

The agents in the model are individual basking sharks. The model is run with 200 sharks because, while the largest sighting in Inishowen was 60 sharks, the largest sighting in Ireland was 160 sharks. The sharks are not distinguishable by size, age, or sex due to a lack of available data on population demographics. Sharks do not give birth or die in the model. The patches within the model can be land or water. The water patches are each assigned a unique number of zooplankton. In each 24 h cycle, the total amount of zooplankton in the model is calculated from the CPR data.

Zooplankton: source data and spatial distribution in model

Zooplankton data came from a formal data request to the Sir Alister Hardy Foundation for Ocean Science (SAHFOS), which conducts the CPR Survey (Johns 2020). These data indicate absolute counts of individual zooplankton for a number of species, collected from synoptic sampling across shipping transects (Richardson et al. 2006). Daily CPR data were requested for this research, with the daily abundance totaled by species (Johns 2020, doi: 10.17031/64d23cc3a1069). For input into the model, the data were separated into two distinct categories: (1) Calanus species (“Cal”) and (2) Pseudocalanus2 and Centropages typicus (“Other_ZP”), as basking sharks are known to prefer Calanus and other large copepod species (Sims et al. 2003; Sims and Reid 2002b). Of 7,918 dates, 6,264 (79%) had no CPR samples; linear interpolation was performed to fill the data gaps. Linear interpolation was chosen to preserve the overall trajectory and magnitude of change in zooplankton data while limiting assumptions about seasonality. To compare the interpolated model zooplankton data with the CPR data, the monthly averages of the CPR data and the interpolated data were both normalized to fall between 0 and 1 using min-max normalization; the ME, RMSE, and MAE were calculated (Suppl. material 3; Table 1). The results indicate that the interpolated zooplankton data input into the model do not show a strong fit with the CPR data. Although this is a limitation, the CPR data were retained because they are the only widely available long-term dataset. They were used to demonstrate the proof of concept for this modeling framework but can be replaced in future iterations as higher-resolution or improved datasets become available.

Table 1.

Parameters that can be modified.

Parameter Explanation Setting
Threshold_ZP Minimum amount of zooplankton (Cal and Other_ZP combined) required for a shark to stay in or move to a patch. Count of individual zooplankton. 0–1000000000000
No_Eat_Min Number of days a shark must encounter a patch with less than the Threshold_ZP before leaving the model. 0–100
Sense-Distance How far a shark can see, smell, or hear (in km). This is doubled when sharks are locating other sharks. 0–100
Swim-Speed The distance a shark can swim (in km). 0–100
Return-Season Number of days it will take a shark to return after it has left in response to reaching the No_Eat_Min. 0–100
Cal_% Percentage of patches with Calanus copepods. 0–100
Other_ZP_% Percentage of patches with Pseudocalanus spp. and Centropages typicus. 0–100
Friend_Min Number of other sharks a patch must have to attract a shark. 0–100

At each time step, the average abundance of zooplankton across all water patches for the entire model environment is assumed to be given by the linearly interpolated CPR data. Since the CPR samples 3 m3 of water, the values are divided by three to obtain an average value per 1 m3. To distribute the zooplankton across the area, it is first assumed that only a percentage of patches contain each category of zooplankton on a given day; these percentages can be altered with a slider using the parameters Cal_% and Other_ZP_% (see Table 1). The average abundance from the CPR for each category is thus divided by the percentage of patches that have zooplankton to calculate the average number of zooplankton in the populated patches. To add spatial variability, the number of zooplankton in these patches is assumed to be normally distributed around this average with a standard deviation of 0.1. Simulated CPR sampling was also conducted in model runs to mimic CPR data collection, as recommended by CPR (Richardson et al. 2006). ME, RMSE, and MAE comparisons of the in-model zooplankton sampling indicated that the simulated zooplankton samples are representative of the CPR data (Suppl. material 3, Table 2).

Table 2.

Settings for Test A and Test B.

Parameter Test A: Food/Social Test B: Food
Threshold_ZP 3.00E+12 1E+11
Sense-Distance 10 10
Swim-Speed 9 9
Cal_% 17 10
Other_ZP_% 17 10
Friend_Min 5 5
No_Eat_Min 14 4
Return-Season 20 20

Assumptions in basking shark behavior/movement

The model assumes annual fidelity for all simulated sharks (200 out of an estimated 10,000 worldwide (Hoelzel et al. 2006)) at the start of the season. Basking sharks have been documented to exhibit site fidelity in the model area, as well as broadly in Ireland and Scotland (Johnston et al. 2019; Thorburn et al. 2024; McInturf et al. 2026). Basking sharks are also consistently sighted in the model region, following seasonal patterns (Jabado et al. 2025; McInturf et al. 2026). Each year, sharks are randomly assigned a migration date from 1 April–31 May to more closely simulate the inconsistent arrival patterns of basking sharks along coastlines in the North Atlantic (Doherty et al. 2017a; McInturf et al. 2026). Each shark enters the model area at a random location along the “watery” edges of the model area – east, west, and north. This is randomized because, while the general trend appears to be that sharks move north to south along the Irish coast, this pattern is not 100% consistent, and sharks may also move from Scotland to Ireland (Doherty et al. 2017a; Johnston et al. 2019; McInturf et al. 2026).

Within the model, the senses of sight, smell, and hearing are collectively called “Sense-Distance.” This is not a fixed parameter but can be altered before the start of the model run, as the mechanism by which basking sharks sense zooplankton is still unclear3. When seeking other sharks, a shark’s Sense-Distance is doubled, under the assumption that basking sharks can either hear, see, or smell other basking sharks at a greater distance than they can zooplankton, especially in light of research indicating that breaching serves as a form of long-range signaling between basking sharks (Harwood 2001; Gore et al. 2018; Chapuis et al. 2019; Berrow et al. 2025; Klöcker et al. 2026). The distance that sharks can sense remains static throughout the model run, as the model does not account for ocean currents, wind, or other factors influencing the distance over which sound and smell travel through water, or how turbidity can affect sight, to reduce unnecessary complexity. While ocean dynamics can affect smell and sound, these were omitted from the model for simplicity to reduce environmental complexity and maintain focus on the target behavior of the sharks.

Sharks retain a memory of high-zooplankton patches (defined as those that contain three times the threshold level of zooplankton, given by the parameter “Threshold_ZP”) to simulate intermittent site fidelity (demonstrated by field observations), which has been attributed to resource availability (Doherty et al. 2017b; Austin et al. 2019). While the precise mechanism driving site fidelity is still under debate, this behavior is associated with long-term memory in other migratory species, and laboratory experiments have documented the ability of sharks to remember information related to food and even social groups over extended periods (Anderson et al. 2011; Fuss and Schluessel 2015; Marosi et al. 2026).

During each daily time step, each basking shark first decides whether to move to a new patch. This decision is based on two factors: (1) the amount of zooplankton and (2) the number of sharks within its current patch. Sharks determine whether there is sufficient zooplankton to meet the threshold level of zooplankton (Threshold_ZP, which can be altered via a slider) for the total number of sharks in the patch (Fig. 2). If there is sufficient zooplankton to meet the threshold level for each shark, the shark will remain in the patch and not select a new patch to move to. If there is insufficient zooplankton to meet the threshold level for each shark, the model assumes that the shark seeks a new patch to move to, a behavior basking sharks have been observed exhibiting when zooplankton is reduced (Sims and Quayle 1998; Sims 1999; Sims et al. 2003). Other filter-feeding shark species have been documented to change behavior based on intraspecific competition (Montero-Quintana et al. 2021).

Figure 2. 

Shark decision pathway for Food, Social, and Food/Social submodels. Shark decision pathway under different submodels. Each day, sharks complete this decision tree. Note that if a patch that meets the condition is identified, under all versions, sharks take the following action: if it is within swimming distance, move to it; if it is beyond swimming distance, swim toward it. The Random control is not included in this figure, as sharks simply select a random patch to move to each day and do not follow any decision pathway related to either food availability or the location of other sharks.

Where and why a shark will move next depend, in part, on the submodel chosen (see Fig. 2 and Section Submodels for more details on the submodels). In general, sharks will select a list of patches within their Sense-Distance that meet the submodel requirements. Sharks then sort the patches in descending order of desirability, as prescribed by the submodel. The first patch in the list becomes the shark’s target patch. To avoid land, sharks assess whether there is a patch of land between them and the target patch. If there is land in their linear travel route, the next most desirable patch becomes their target patch. They repeat this process until they find a target patch for which there is no intervening land. In the Food and Food/Social submodels, if sharks cannot find a target patch in this way, they will review their memory4 of high-zooplankton patches and set the closest patch that previously had high levels of zooplankton as their target patch. They will follow the same procedure to avoid land.

Assuming a target patch is ultimately found, sharks then determine whether the target patch is within the distance they can swim in one time step (given by the parameter Swim-Speed), which may be smaller than Sense-Distance. If the target patch is within the distance set by Swim-Speed, the shark will move to that patch in a single time step. If it is farther away, the shark will move the maximum distance allowed by the Swim-Speed parameter toward the target patch. Sharks do not automatically continue toward the selected patch after 24 h, as it is assumed that sharks alter their behavior based on new information and always preferentially select closer patches over those that are farther away. If the sharks cannot find a target patch, they will swim in a random direction, following the same procedure to avoid land, for the distance given by Swim-Speed. This overall process is repeated at each time step, even if a shark moved only partway toward its target patch in the previous time step.

At each time step, sharks determine whether there was sufficient zooplankton (both Cal and Other_ZP) to support the shark (i.e., meet the minimum level set by Threshold_ZP). If not, the shark records this as a day on which it did not ‘eat’.5 If a shark has not eaten for the minimum number of days given by the parameter No_Eat_Min, it either dives to deeper water, where it may find deeper patches of zooplankton (Elcock et al. 2026), or migrates out of the model area entirely, a behavior documented in the North Atlantic (Sims et al. 2000; Cotton et al. 2005; Doherty et al. 2017b; McInturf et al. 2026)6. Sharks will “migrate” back into the model area after a period of time given by the parameter “Return-Season,” as it is well documented that sharks return to surface waters in Ireland later in the season (Sims and Reid 2002a; McInturf et al. 2026). They will again enter the model from the north, east, or west, with the specific location at the edge of the model area chosen at random, as basking sharks do not follow universal migration or sighting patterns in the North Atlantic (Sims and Reid 2002a; Doherty et al. 2017a; McInturf et al. 2026). Sharks will then resume normal decision-making, retaining their memory of high-zooplankton patches from year to year.

At each time step, the following data are collected: (1) the total number of basking sharks within the model area; (2) for patches containing two or more basking sharks, the number of sharks, the amount of zooplankton in the patch, and the location of the patch; and (3) Total Aggregations, i.e., the number of patches with two or more basking sharks. The model also includes randomized sampling of patches (referred to as Pseudo-Sighting Reports), intended to simulate the IBSG/IWDG public sightings reports by hikers and boaters who frequent the area in and around the Inishowen Peninsula (Gray et al. 2022). Twenty patches of water are randomly sampled every day, and if there is a shark in the patch, the model records the number and location of the sharks. Total Aggregations will only record two or more sharks, but Pseudo-Sighting Reports record any shark sighting, in accordance with IBSG/IWDG sightings. Both Pseudo-Sighting Reports and Total Aggregations outputs also record the latitude, longitude, and zooplankton level of each patch where the sharks were seen. Finally, simulated sampling of model zooplankton is conducted. Ten random patches are sampled every day, and the number of zooplankton is averaged across these patches. This sampling was conducted as a rough simulation of CPR sampling, as recommended by the CPR (Richardson et al. 2006).

Shark decision pathway under different submodels. Each day, sharks complete this decision tree. Note that if a patch that meets the condition is identified, under all versions, sharks take the following action: if it is within swimming distance, move to it; if it is beyond swimming distance, swim toward it. The Random control is not included in this figure, as sharks simply select a random patch to move to each day and do not follow any decision pathway related to either food availability or the location of other sharks.

Submodels

While all non-behavioral elements of the model remain the same, the model consists of four alternative submodels of shark behavior related to their decisions about target patches once they have decided to move from their current patch (Fig. 2):

  1. Social: Sharks search only for other sharks. They judge patches within their Sense-Distance by the number of sharks. They select target patches using the parameter “Friend_Min,” which defines the minimum number of sharks that must be in a patch, and then sort the target patches by the number of sharks. If they are unable to locate any other patches with sharks, they will follow the Random submodel. They will still leave an area if there is insufficient zooplankton in a target patch to support the number of sharks 7
  1. and will still migrate out of the area if they have not eaten after the set number of days. Sharks do not utilize memory of high-zooplankton patches.
  2. Food: Sharks search only for food (zooplankton). Sharks judge patches within their Sense-Distance by whether the patches meet the Threshold_ZP parameter based on the combined abundance of Cal and Other_ZP, and then sort the patches based on the amount of Cal. The sharks also retain a memory of previous high-zooplankton patches outside their Sense-Distance and will consider these during times of overall low zooplankton. If they are unable to locate any other patches with food, they will move following the Random submodel. Again, they will still leave an area if there is insufficient zooplankton in a target patch to support the number of sharks and will still migrate out of the area if they have not eaten after the set number of days.
  3. Food/Social: This is a combination of the Food and Social submodels, and sharks execute all decision-making from both submodels. Sharks first search for food, then other sharks 8
  1. within their Sense-Distance. If they cannot find a patch that meets the preferred requirements within this range, they will select a patch from their memory of high-zooplankton patches. If they are unable to locate any other patches with food or sharks, they will move following the Random submodel. Again, they will still leave an area if there is insufficient zooplankton in a target patch to support the number of sharks and will still migrate out of the area if they have not eaten after the set number of days.
  2. Random: Sharks swim in a random direction for the distance given by Swim-Speed, while avoiding land. This serves as a control. They will still migrate out of the area if they have not eaten after the set number of days.

Data analysis

Only IBSG/IWDG sightings from 1982–2018, from April through October, were included in the data analysis, with the exception that sightings in the last week of March were included in the April count to increase the dataset available for data analysis.9

To compare the model output data with the real-world sightings data, the average number of sharks per aggregation for each month (hereafter referred to as monthly aggregation size) was determined for the Total Aggregations, Pseudo-Sighting Reports, and IBSG/IWDG data. The average monthly number of aggregations was also calculated for all datasets.

All datasets were normalized using minimum–maximum normalization, and the mean error (ME), mean absolute error (MAE), and root mean square error (RMSE) were then calculated between the Total Aggregations and IBSG/IWDG monthly averages and between the Pseudo-Sighting Reports and IBSG/IWDG monthly averages. RMSE, ME, and MAE were used because they are statistical measures used for model forecasting of non-normal and inconsistently patterned datasets (Hodson 2022). ME, RMSE, and MAE are used in time-series forecasting and are less forgiving of large errors than small ones, while RMSE is specifically sensitive to outliers. These measures are well suited for time series with no clear trend but clear seasonality.

The normalized average monthly aggregation sizes were also graphed as time series and box plots. Visual comparisons were then made between model outputs and the IBSG/IWDG sightings data. Each model run was compared with two subsets of the IBSG/IWDG sightings data: the full Irish dataset (referred to as “Ireland”) and the subset of sightings reported from within the simulated model area (referred to as “Inishowen”).

Exploratory testing

Initial exploratory testing to determine appropriate model parameter settings was largely qualitative, with initial parameter values drawn from the existing literature for Threshold_ZP (Sims 1999, 2008), Swim-Speed (Skomal et al. 2009; Doherty et al. 2017a; Johnston et al. 2018), and Sense-Distance (Sims 2008; Collin 2012; Dove and Pierce 2021), though many of these are still under debate. No or limited literature exists for the No_Eat_Min, Return-Season, Cal_%, Other_ZP_%, or Friend_Min settings, so experimentation was required to determine the settings that achieved the most realistic results. Parameters were initially modified on an individual basis and then in gradual combination.

Small experimental trials were run under different setting combinations and compared with a predefined definition of realism derived from the IBSG/IWDG data. The majority of basking shark aggregations in Ireland range from 2–10 sharks, with aggregations of 15–40 sharks occurring approximately 15% of the time and a maximum aggregation size of at least 60 but no greater than 160. A range of semi-realistic settings was identified by comparing the aggregation size distribution within the model with the IBSG/IWDG data, an approach consistent with pattern-oriented modeling (Railsback and Grimm 2019). A systematic sweep was not feasible given the model’s computational demands, so an iterative approach was used to identify the parameters that met the realism criteria.

The majority of these experimental trials resulted in unrealistic frequencies of aggregation sizes, with aggregations often not exceeding 20 sharks or dramatically exceeding 160 sharks. Experimentation continued in an ad hoc manner until model results with relatively realistic maximum aggregation sizes and frequencies of aggregation sizes were achieved. Approximately 2,000 model tests were run during this exploratory process.

The results of the exploratory tests that produced realistic aggregation size distributions and realistic maximum aggregation sizes were then compared with the IBSG/IWDG data using RMSE, ME, and MAE analyses. Those tests with the lowest RMSE results when compared with the IBSG/IWDG data10 were selected for further testing to identify the parameters that achieved the most realistic results. RMSE was given more weight because it is less sensitive to outliers than ME and MAE, though consistency across all three statistical results was considered a factor.

Parameter selection

Once settings that produced strong RMSE results were identified, more detailed preliminary testing was conducted to identify the settings that produced the most realistic results for all submodels. Initial calibration testing showed high consistency across trials for the Random and Social submodels but low similarity with the IBSG/IWDG data. This is likely because the Random and Social submodels rarely produced aggregations of more than two or three sharks under any combination of settings. Thus, preliminary tests were conducted using only the Food and Food/Social submodels to save computational power and time. To further reduce computational needs, parameter selection tests were conducted with a maximum of 100 sharks in the model. The reduction from 200 to 100 basking sharks during parameter selection was necessary for computational feasibility. Reducing the number of agents during these exploratory stages allowed a broader range of parameter combinations to be screened within the available computational budget, with the full 200-shark, 50-trial testing reserved for confirmatory runs of the final selected settings.

Small trials (5–10) were then run, and the RMSE, ME, and MAE were calculated until six settings emerged as producing the most realistic output (Suppl. material 3, Tables 3, 4). A further 10 trials of these six settings were run, and the output across all trials was averaged. The results of the RMSE, ME, and MAE indicated which settings were the most comparable to the IBSG/IWDG data. When comparing Total Aggregations with either Inishowen or Ireland, the setting that resulted in the lowest RMSE result for any submodel was selected as the recommended “default” setting for the Food and Food/Social submodels (Suppl. material 3, Table 3). These default settings received relatively consistent RMSE, ME, and MAE results when comparing data with both Ireland and Inishowen.

Table 3.

Average number of shark aggregations for 50 trials of Test A and Test B.

Submodel Total Aggregations Pseudo-Sighting Reports
Test A (best fit for Food/Social submodel) Food/Social 4099 208
Social 568 170
Random 629 172
Test B (best fit for Food submodel) Food 73791 322
Social 105 178
Random 173 170
Table 4.

Comparison of the average number of aggregations per month with IBSG/IWDG data.

Submodel Inishowen All of Ireland
ME RMSE MAE ME RMSE MAE
Test A (best fit for Food/Social submodel) Total Aggregations Food/Social 0.03 0.16 0.06 0.0 0.17 0.08
Social 0.27 0.39 0.28 0.25 0.36 0.26
Random 0.16 0.24 0.17 0.13 0.21 0.16
Pseudo-Sighting Reports Food/Social 0.11 0.18 0.13 0.06 0.19 0.13
Social 0.23 0.28 0.24 0.18 0.27 0.22
Random 0.27 0.31 0.27 0.22 0.29 0.25
Test B (best fit for Food submodel) Total Aggregations Food 0.46 0.51 0.47 0.43 0.49 0.45
Social 0.10 0.18 0.12 0.07 0.17 0.11
Random 0.14 0.24 0.16 0.11 0.22 0.15
Pseudo-Sighting Reports Food 0.17 0.21 0.18 0.14 0.21 0.18
Social 0.25 0.32 0.26 0.22 0.29 0.23
Random 0.30 0.37 0.31 0.28 0.35 0.29

Preliminary testing identified Test A as the best parameter setting for the Food/Social submodel, while Test B was identified as the best parameter setting for the Food submodel. Fifty trials were run under these parameter settings, with the output compared with the IBSG/IWDG sightings data.

Full-scale simulation runs

Once the settings that produced the most realistic outputs for the Food/Social (Test A) and Food (Test B) submodels were identified, 50 trials were run for each setting (Table 2). A maximum of 200 sharks were included in these final simulations, and the RMSE, ME, and MAE were calculated for the average monthly size of aggregations and the average monthly number of aggregations for both the Total Aggregations and Pseudo-Sighting Reports. The Social submodel was included in Tests A and B for comparison, but no calibration test resulted in realistic aggregations, so tests were not run independently on this submodel. The Random submodel was included in both tests as a control.

Normalized results from Tests A and B were plotted as box plots and time-series graphs alongside normalized results from Inishowen and Ireland. The average size of shark aggregations per day (derived from Total Aggregations) was graphed in a scatter plot against the average amount of zooplankton in the same patches on a log scale.

Sensitivity and robustness analysis

Sensitivity and robustness analysis (SA/RA) testing (Railsback and Grimm 2019) was then conducted using the settings in Test A for the Food/Social submodel, as this is the most ecologically complete submodel (Table 2). All tests were conducted with a maximum of 200 sharks in the model area, and the average number of aggregations and average maximum aggregation size were calculated for sets of 10 trials.

Sensitivity analyses were conducted by altering the settings of Test A one by one, each by ±5%, ±35%, and ±70%, as described by Railsback and Grimm (2019). The change in parameter settings was calculated following the formula S+ = (C+ − C) / (dp/p), where C is the result for Test A and C+ is the result for the sensitivity test. p refers to the baseline parameter setting, while dp is the percent change in the parameter. The difference (S+) in the average number of aggregations and average maximum aggregation size was calculated for each parameter.

Robustness testing was conducted by removing, one at a time, the following five behavioral mechanisms from the Food/Social submodel: sharks did not check food availability in their patch; sharks did not seek other sharks; sharks did not “migrate” out of the model area when they had not eaten after a set number of days; sharks did not maintain a memory of high-zooplankton patches; and sharks did not seek patches of high zooplankton. The percent difference in the average number of aggregations and average maximum aggregation size was calculated between the robustness analysis and the results of Test A.

Results

Exploratory testing and parameter selection

Exploratory testing narrowed the range of parameter settings that produced semi-realistic results for the Food and Food/Social submodels. No tests produced realistic results for either the Social or Random submodels. From the exploratory testing, it was determined that setting Sense-Distance to 10, Swim-Speed to 9, Friend_Min to 5, and Return-Season to 20 produced the most realistic results for both the Food and Food/Social submodels (Table 2). However, different settings were identified for Threshold_ZP (3 × 1012 for the Food/Social submodel and 1 × 1011 for the Food submodel), Cal_%, and Other_ZP_% (17% for both parameters in the Food/Social submodel and 10% for both parameters in the Food submodel). These combinations of parameter settings resulted in notably lower RMSE, ME, and MAE results than any of those in the preliminary tests (Suppl. material 3, Tables 3, 4). These two settings were then selected for full-scale simulation runs.

Full-scale simulation runs

The number of Total Aggregations in Test A (best fit for the Food/Social submodel) ranged from 1,712–3,925 aggregations for the Food submodel but had a lower range for the Social (143–568 aggregations) and Random submodels (165–629 aggregations; Table 3). In Test B (best fit for the Food submodel), the number of Total Aggregations ranged from 72,484–74,914. Again, the number of Total Aggregations had a much lower range for the Social (81–124 aggregations) and Random submodels (82–253 aggregations; Table 3).

When comparing the normalized number of aggregations per month (averaged across all trials) from the model tests with the IBSG/IWDG data, the Food/Social submodel produced the lowest ME, RMSE, and MAE results (0.0–0.17) of any submodel in either Test A or Test B (Table 4). While the Total Aggregations of the Food submodel in Test B did not produce particularly realistic results (ranging from 0.43–0.51), the Pseudo-Sighting Reports produced results comparable to the Food/Social submodel in Test A (0.14–0.29). The Social and Random submodels in Test A resulted in higher ME, RMSE, and MAE values than the Food/Social submodel (between 0.16 and 0.21); however, they received notably lower ME, RMSE, and MAE values in Test B, though with a wider range (0.10–0.35; Table 4).

A comparison of the time series of the normalized number of aggregations per month (averaged across all trials) for both Test A and Test B shows that the Food/Social submodel in Test A produced the most realistic time-series output, although it was not a perfect match to the IBSG/IWDG data. However, the Food/Social submodel produced more episodic, irregular increases and decreases in the number of aggregations, while the Food submodel produced more rhythmic or seasonal cycling that was less comparable to the IBSG/IWDG data. This was exhibited by both the Total Aggregations and Pseudo-Sighting Reports. When considering the box plots, only the Food/Social submodel (Test A) produced an interquartile range that overlapped with data from both Ireland and Inishowen, although the normalized Total Aggregations displayed a more comparable range than the Pseudo-Sighting Reports (Fig. 6). The Food submodel (Test B) showed no interquartile overlap with the IBSG/IWDG data (Fig. 7).

Total number of aggregations throughout the entire model run (1982–2018) for Test A and Test B. Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations count only groups of two or more sharks but count all aggregations in the model each day. See Table 2 for parameter settings for Tests A and B (Figs 3, 4, 5).

Figure 3. 

Average number of aggregations per month for Test A. Comparison of the average number of aggregations per month throughout the model run (1982–2018), averaged across 50 trials for Test A (see Table 2 for parameter settings). Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations count only groups of two or more sharks but count all aggregations in the model each day. Data were normalized via min-max normalization and graphed on a time-series plot alongside the model output for both Ireland (all IBSG/IWDG data) and Inishowen (the modeled area; see Fig. 1). The green line represents the average daily total zooplankton input into the model area, averaged per month and normalized via min-max normalization.

Figure 4. 

Average number of aggregations per month for Test B. Comparison of the average number of aggregations per month throughout the model run (1982–2018), averaged across 50 trials for Test B (see Table 2 for parameter settings). Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations count only groups of two or more sharks but count all aggregations in the model each day. Data were normalized via min-max normalization and graphed on a time-series plot alongside the model output for both Ireland (all IBSG/IWDG data) and Inishowen (the modeled area; see Fig. 1). The green line represents the average daily total zooplankton input into the model area, averaged per month and normalized via min-max normalization.

Figure 5. 

Box plot of the average number of aggregations and average aggregation size per month for Test A. Box plots comparing the average number of aggregations and average size of aggregations per month throughout the model run (1982–2018), averaged across 50 trials for Test A (see Table 2 for parameter settings). Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations only count groups of two or more sharks but count all aggregations in the model each day. Data were normalized via min-max normalization and graphed as a box plot series alongside the model output for both all of Ireland (all IBSG/IWDG data) and Inishowen (the modeled area; see Fig. 1). Boxes denote the upper and lower quartiles; the solid line marks the median, and the dotted line marks the mean; the uppermost and lowermost points indicate the maximum and minimum values.

Figure 6. 

Box plot of average number of aggregations and average size per month for Test B. Box plots comparing the average number of aggregations and average size of aggregations per month throughout the model run (1982–2018), averaged across 50 trials for Test B (see Table 2 for parameter settings). Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations only count groups of two or more sharks but count all aggregations in the model each day. Data were normalized via min-max normalization and graphed as a box plot series alongside the model output for both all of Ireland (all IBSG/IWDG data) and Inishowen (the modeled area; see Fig. 1). Boxes denote the upper and lower quartiles; the solid line marks the median, and the dotted line marks the mean; the uppermost and lowermost points indicate the maximum and minimum values.

Figure 7. 

Average size of aggregations per month for Test A. Comparison of the average size of aggregations per month throughout the model run (1982–2018), averaged across 50 trials for Test A (see Table 2 for parameter settings). Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations only count groups of two or more sharks but count all aggregations in the model each day. Data were normalized via min-max normalization and graphed on a time-series plot alongside the model output for both all of Ireland (all IBSG/IWDG data) and Inishowen (the modeled area; see Fig. 1). The green line represents the average daily total zooplankton input into the model area, averaged per month and normalized via min-max normalization.

Comparison of the average number of aggregations per month, averaged across 50 trials for Tests A and B (see Table 2 for parameter settings). Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations count only groups of two or more sharks but count all aggregations in the model each day. Data were normalized via min-max normalization, and the ME, RMSE, and MAE were calculated by comparing the model output with both Ireland (all IBSG/IWDG data) and Inishowen (the modeled area; see Fig. 1).

When comparing the normalized monthly aggregation size (averaged across all trials) across all tests, the Food/Social submodel (Test A) resulted in the lowest RMSE value (0.18; Table 5). The next lowest RMSE value was for the Food submodel in Test B (0.21; Table 5). Similarly, the Food/Social submodel in Test A resulted in the lowest ME and MAE values, with the Food submodel in Test B having the second-lowest ME and MAE values (Table 5). This trend was repeated with the Pseudo-Sighting Reports for both Test A and Test B, with the Food/Social submodel having the lowest RMSE, ME, and MAE values (Table 5). In all tests, the Social and Random submodels produced very high ME, RMSE, and MAE values (Table 5). This indicates that the Food/Social submodel under the settings of Test A produces the most realistic model output.

Table 5.

Comparison of average aggregation size per month with IBSG/IWDG data.

Submodel Inishowen All of Ireland
ME RMSE MAE ME RMSE MAE
Test A (best fit for Food/Social submodel) Total Aggregations Food/Social 0.10 0.18 0.13 0.08 0.18 0.12
Social 0.62 0.70 0.63 0.61 0.68 0.61
Random 0.58 0.67 0.58 0.56 0.65 0.57
Pseudo-Sighting Reports Food/Social 0.07 0.16 0.11 0.05 0.15 0.10
Social 0.38 0.43 0.39 0.36 0.41 0.37
Random 0.43 0.48 0.44 0.41 0.45 0.42
Test B (best fit for Food submodel) Total Aggregations Food 0.17 0.21 0.18 0.14 0.21 0.18
Social 0.25 0.32 0.26 0.22 0.29 0.23
Random 0.30 0.37 0.31 0.28 0.35 0.29
Pseudo-Sighting Reports Food 0.16 0.27 0.19 0.14 0.26 0.18
Social 0.36 0.43 0.37 0.34 0.40 0.35
Random 0.35 0.40 0.40 0.33 0.38 0.34

The Food/Social submodel in Test A displayed the strongest qualitative similarities to the timing of the IBSG/IWDG data, following a nonuniform but seasonal pattern of aggregation size through time. However, increases in the average aggregation size occurred in the model with a slight lag compared with the IBSG/IWDG data (Fig. 7). The Food submodel in Test B also produced a semi-realistic output, with a notably higher number of larger aggregations than documented in the IBSG/IWDG data (Fig. 8). The Social and Random submodels produced unrealistic time-series outputs in both Tests A and B (Figs 7, 8). The Pseudo-Sighting Reports for both the Food and Food/Social submodels depicted a similar delayed temporal trend in average aggregation size per month compared with the IBSG/IWDG data (Fig. 6). The Food and Food/Social Pseudo-Sighting Reports still produced a more realistic time series than the Social and Random submodels. When considering the box plots of average monthly aggregation size, the Food/Social submodel (Test A) was a better fit to the IBSG/IWDG data (Fig. 5). Within Test B, the Food submodel’s aggregation size showed a slightly closer fit than its average monthly number of aggregations (Fig. 6).

Figure 8. 

Average size of aggregations per month for Test B. Comparison of the average size of aggregations per month throughout the model run (1982–2018), averaged across 50 trials for Test B (see Table 2 for parameter settings). Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations only count groups of two or more sharks but count all aggregations in the model each day. Data were normalized via min-max normalization and graphed on a time-series plot alongside the model output for both all of Ireland (all IBSG/IWDG data) and Inishowen (the modeled area; see Fig. 1). The green line represents the average daily total zooplankton input into the model area, averaged per month and normalized via min-max normalization.

Graphing the average daily aggregation size against the average amount of zooplankton in the same patches showed that zooplankton availability had no impact on aggregation size in the Social or Random submodels. Under the settings of Test A, the Food/Social submodel showed a slight trend, with the majority of aggregations being smaller across a wider range of zooplankton availability (Fig. 9). A similar trend was seen in the Food submodel (Test B), but with a higher number of aggregations and larger aggregations. In both, when aggregation sizes were larger, they correlated with an increased level of zooplankton (Fig. 9), a trend also shown in the time-series plots.

Figure 9. 

Daily aggregation size compared with zooplankton abundance in patches with aggregations.

Comparison of the average size of aggregations per month, averaged across 50 trials for Tests A and B (see Table 2 for parameter settings). Pseudo-Sighting Reports include any shark “sighted” during a random sample of 20 patches, including single sharks, while Total Aggregations only count groups of two or more sharks but count all aggregations in the model each day. Data were normalized via min-max normalization, and the ME, RMSE, and MAE were calculated by comparing the model output with both all of Ireland (all IBSG/IWDG data) and Inishowen (the modeled area; see Fig. 1).

The results of 50 trials were averaged together. The scatter plot depicts the average aggregation size per day (taken from Total Aggregations) compared with the average amount of zooplankton in the patches where the aggregations were observed. Zooplankton abundance was graphed on a logarithmic scale.

Sensitivity and robustness analysis

The model parameter with the highest impact on the number of aggregations in the Food/Social submodel was Cal_%, with Return-Season having the strongest impact on the size of the aggregations. Return-Season and Threshold_ZP also had a strong impact on the number of aggregations, with No_Eat_Min having an impact on the size of aggregations (Table 6).

Table 6.

Sensitivity testing of the Food/Social submodel.

Average number of aggregations
Parameter –70% –35% –5% +5% +35% 70%
Cal_% –7.36 –7.36 –2.10 –3.09 –1.34 –1.00
Friend_Min –0.74 –0.74 –0.42 –0.17 –0.05 –0.07
No_Eat_Min 0.30 0.30 1.03 –0.49 1.09 0.50
Other_ZP_% –0.41 –0.41 0.19 –0.23 –0.03 –0.01
Return-Season –2.69 –2.69 –1.99 –6.84 –1.23 –0.71
Sense-Distance 0.79 0.79 –0.13 –1.26 –0.9 –0.70
Swim-Speed 1.14 1.14 1.65 0.77 0.22 0.15
Threshold_ZP –6.18 –6.18 –1.30 –3.50 –1.36 –0.94
Average maximum aggregation size
Cal_% –2.46 –1.73 –0.75 0.17 –0.26 –0.74
Friend_Min 0.01 -0.18 0.07 –0.06 –0.09 –0.14
No_Eat_Min 0.46 -0.82 –1.89 –1.75 1.97 0.89
Other_ZP_% 0.00 -0.23 –0.96 1.59 0.01 –0.05
Return-Season –0.47 0.13 –5.95 –5.50 –1.17 –0.58
Sense-Distance 1.20 1.28 0.42 –0.93 –0.15 –0.13
Swim-Speed 1.27 1.42 1.48 0.93 0.42 0.16
Threshold_ZP –2.23 –1.77 1.28 1.52 0.2 –0.27

Removing No_Eat_Min had the most significant impact on both the number and size of aggregations (Table 6). Removing the behavior of basking sharks seeking other sharks strongly impacted the size of aggregations but not the number of aggregations. Removing memory had the lowest impact on both the number and size of aggregations (Fig. 10). Removing the behavior whereby basking sharks assess food availability in their patch also had a strong impact on aggregation size. All removals except those of memory and migration resulted in an overall decrease in the number and size of aggregations.

Figure 10. 

Robustness analysis. Robustness testing was conducted on the Food/Social submodel with the parameter settings of Test A. The listed behaviors were removed one at a time, and 10 trials were run for each behavior that was removed. The results of 10 trials were averaged. The percent difference in the average number of aggregations and average maximum aggregation size was calculated between the robustness analysis and the results of Test A. The 95% confidence intervals were calculated using a t-distribution (df = 9) across the 10 trials.

Discussion

This model demonstrates the viability of IBMs as a mechanism for modeling basking shark aggregations. The results of the different submodels and the robustness analysis make clear that the behavior of the simulated sharks has a clear impact on the model output, demonstrating that aggregations were not a result of random chance but of shark (agent) behavior.

While the Social and Random submodels had higher consistency between trials, this is because both models rarely produced aggregations of more than two sharks. In the Social submodel, basking sharks aggregate based on conspecific cues alone, without assessing food availability in a patch. Because there is no shared environmental cue drawing multiple individuals toward the same location, sharks continuously reassess and move toward whichever nearby individuals satisfy their social criteria, but with no environmental factors causing multiple sharks to converge on or near the same area. The fact that this submodel rarely produces aggregations larger than two individuals suggests that social attraction in isolation is insufficient to generate the larger aggregations observed in the IBSG/IWDG data.

In contrast, the more ecologically complete model (Food/Social submodel) resulted in lower RMSE, ME, and MAE results and more realistic time series compared with the Social and Random submodels. The submodel that produced results that were most statistically and qualitatively similar to the IBSG/IWDG data was the submodel that included both food availability and social interactions as drivers of basking shark behavior, resulting in the emergence of aggregations not unlike those observed in the wild. This is consistent with research indicating that basking shark aggregations are the result of a combination of food availability and social interactions (Crowe et al. 2018; Lieber et al. 2020; Sims et al. 2022).

The removal of certain behaviors in the robustness analysis had a significant impact on the model outputs. Removing the behavior of seeking other sharks increased both the number and size of the aggregations, as demonstrated by the robustness tests and the results of the Food submodel in Test B. Notably, the Food submodel resulted in trends that were a weaker fit to the IBSG/IWDG data, indicating that food availability is insufficient to explain these aggregations.

Migration from the surface waters of the model area clearly impacted the results of the model, as demonstrated by the results of the robustness tests. Removing this behavior greatly inflated the number of aggregations, as 200 sharks remained in the model area for the entirety of each modeled year, an unrealistic behavior for sharks in this area (McInturf et al. 2026). While the time it takes a shark to either leave an area or return to it is not known, the 14-day No_Eat_Min and 20-day “Return-Season” result in annual seasonal patterns that match the IBSG/IWDG sightings reports, which show a spike in sightings in April, followed by a delay in sightings for 1–3 months and another increase in reports during the autumn months (McInturf et al. 2026). This likely correlates with seasonal boom-and-bust cycles of zooplankton (Sims and Quayle 1998; Folt and Burns 1999; Sims and Reid 2002b; Sims et al. 2003; Cotton et al. 2005; Feng et al. 2014; Hoogenboom et al. 2015; McInturf et al. 2026). While the reasons sharks either move to deeper water or migrate to or from an area are not known with certainty, the assumptions in the model have produced a reasonable recreation of shark behavior and can be refined as new research becomes available.

When basking sharks did not check a patch for sufficient food to support the sharks in the area, the model resulted in fewer aggregations and smaller aggregation sizes, as sharks did not receive a simulated environmental cue to leave a patch. Similarly, when sharks did not seek out zooplankton, their movements were largely random, resulting in fewer and smaller aggregations. A lack of memory means that the shark is only responding directly to environmental cues and not making any other decisions, something that is unlikely in such a long-lived species that migrates long distances (Anderson et al. 2011; Fuss and Schluessel 2015).

The impact of food availability on aggregations was clearly demonstrated in the model, in agreement with the literature (Berrow and Heardman 1994; Sims et al. 2000, 2022; Cotton et al. 2005; Crowe et al. 2018). Zooplankton availability has been documented as a large-scale driver of basking shark movements in previous research, but its impact on more localized movement is still poorly understood from field studies, making IBMs such as this a valuable tool (Sims et al. 2000; Sims and Reid 2002b). While the Food/Social submodel was most comparable to the IBSG/IWDG data, there were cases in preliminary tests where the Food and Food/Social submodels received similar statistical results, indicating that food availability had a stronger impact on basking shark aggregations, with social impacts being secondary. The results of Test A (best fit for the Food/Social submodel) and Test B (best fit for the Food submodel) indicated that food drives the number of aggregations within the model, while social drivers impact the size of aggregations. The lack of realistic results in the Social submodel, regardless of setting, indicates that a social explanation alone is insufficient to explain the observed aggregation behavior.

Research on the threshold level of zooplankton required for basking sharks to both move to and remain in an area is still limited (Sims et al. 2003). Future research can be used to further inform model development and create a more realistic preyscape. However, this model does demonstrate the viability of developing a semi-realistic preyscape and its applicability in simulating basking shark behavior.

The preliminary tests that had a lower percentage of patches with zooplankton (Cal_% and Other_ZP_% < 20%) resulted in aggregation trends that were more likely to match the IBSG/IWDG sightings data. This indicates that the Inishowen Peninsula may be an area of high zooplankton patchiness and that zooplankton patchiness increases the likelihood of these aggregations. This is further demonstrated by the results of Test B, where only 10% of patches contained zooplankton. Test B had a notably higher number of aggregations in the Food submodel than were documented in the Food/Social submodel (Test A). This is indicative of food availability and the location of food being drivers of aggregations within the model. This may be a reason why the Food/Social submodel performed better than the Food submodel, while in tests with a higher percentage of zooplankton patches, those two submodels were virtually identical (Suppl. material 3, Tables 3, 4). In patchy environments, basking sharks may rely on other sharks as a signal that food is available in a specific location (recall that, in the model, sharks can sense other sharks from a greater distance than they can sense zooplankton). This aligns with research on local enhancement in other shark species (Papastamatiou et al. 2020) and is supported by research on basking sharks, which theorizes that breaching is a method of long-range signaling to conspecifics (Gore et al. 2018; Berrow et al. 2025; Klöcker et al. 2026). In preliminary tests that contained 50% or more patches with zooplankton, the sharks may not have needed to use the presence of other sharks as a proxy for identifying zooplankton. Further research on the real-world distribution of zooplankton in Inishowen should be conducted so that a more realistic preyscape can be included in the model.

Utilizing data from field studies, a rough estimate of the average daily (straight-line) distance traveled by a basking shark was calculated to be approximately 9 km (Skomal et al. 2009), making the Swim-Speed of 9 km (Test A) fairly reasonable. However, new tagging data may continue to be incorporated, and this parameter can be further refined as data become available (Thorburn et al. 2024). Future iterations of the model can also incorporate nonlinear movements, including through the water column, especially in light of new research demonstrating social behavior up to 60 m deep and feeding behavior up to 900 m deep (Rudd et al. 2021a; Elcock et al. 2026). This will allow for simulations of shark behavior beyond surface swimming.

The accuracy of the Sense-Distance parameter is difficult to assess, as the mechanism by which basking sharks sense zooplankton or other sharks is still poorly understood, and the Sense-Distance is likely impacted by the distance a shark can move within a 24 h period. While sharks can sense sound over thousands of meters, they may only be able to sense smells within hundreds of meters (Collin 2012; Dove and Pierce 2021). This distinction may matter for basking sharks, which may seek out zooplankton based on the smell of dimethyl sulfide (DMS) in plankton (Sims and Quayle 1998), though it is not clear how far DMS travels in water and under what conditions (Owen et al. 2021). It is documented that seabirds are able to detect DMS up to 4 km away, and DMS is theorized to be a foraging cue for birds undergoing long migrations (Nevitt and Bonadonna 2005). However, DMS may not be the only environmental cue for basking sharks. Research has documented that whale sharks are attracted to areas of prey by noisy tuna (Dove and Pierce 2021), and basking shark aggregations have been documented at the same time and in the same location as humpback whales (Megaptera novaeangliae), minke whales (Balaenoptera acutorostrata), bluefin tuna (Thunnus thynnus), common dolphins (Delphinus delphis), and seabird species, all of which would generate far-reaching sound (Sims et al. 2022). Recent research has also found that the sound of one basking shark breaching, a common behavior in aggregations, may signal other sharks to breach (Berrow et al. 2025). Because breaching is more likely to occur at night, it is hypothesized to be a method of communication in low-visibility environments (Klöcker et al. 2026). The results of this IBM may lend support to the idea that sound is an important signal of prey or for social signaling in basking sharks, either alongside or in place of other environmental cues, such as DMS. Future iterations of IBM could distinguish sound and smell and perhaps include other species in the model environment. These preliminary findings may also indicate that, as in marine mammals (National Research Council (US) Committee on Potential Impacts of Ambient Noise in the Ocean on Marine Mammals 2003), sound pollution may impact basking sharks’ ability to navigate and find food or mates.

Limitations

This model has several limitations. Most notably, the lack of real-world localized zooplankton data makes it difficult, if not impossible, to assess how realistic the surface environment in the model is for the sharks. Therefore, it is virtually impossible to make a realistic comparison of the location of basking sharks within the model with the IBSG/IWDG sightings data. However, the current model demonstrates that the inclusion of zooplankton is a workable and important component for improving realism when modeling basking sharks. Next steps should involve improved zooplankton data input. Further research could combine satellite phytoplankton data, which can provide fine-scale density and distribution of chlorophyll, with CPR data that provide insight into the abundance and species of zooplankton to simulate zooplankton location and patchiness.

CPR sampling alone does not reflect localized patchiness but samples across transects, so field research intended to assess zooplankton should be conducted in the modeled area to ensure accuracy. New and emerging technology (such as digital holographic cameras and the use of satellite color detection) will also allow future research to obtain a clearer picture of the in situ preyscape for basking sharks (Dyomin et al. 2020; Shunmugapandi et al. 2025). Improved zooplankton data may also allow future model versions to “predict” large aggregations (or at least provide a “high/low” risk estimate) under certain environmental conditions, similar to methods applied to reducing bycatch (Hazen et al. 2018). This could be used for safety warnings for boaters and fishers (i.e., high aggregation risk results in a higher risk of collision).

The memory mechanism is constrained by the current representation of zooplankton in the model. In the model’s current form, zooplankton are randomly distributed each day rather than following a more realistic temporal or spatial distribution. Therefore, the memory mechanism is not able to record a plausible seasonal pattern or long-term trends, thereby limiting its impact on shark foraging behavior. Given this constraint, the ecological significance of memory-based foraging remains largely hypothetical, and the relatively small, though nonzero, effect of removing memory in the robustness analysis should be interpreted with this limitation in mind. Improved zooplankton simulations will allow more thorough testing of the impact of this mechanism on shark behavior.

This research also shows the limitations of publicly reported sightings, as the Pseudo-Sighting Reports in all model versions missed a sizable number of aggregations, even with random sampling of 20 patches per day (a consistent effort not seen in the IBSG/IWDG data). The model data are comparably more systematic than the IBSG/IWDG sighting reports, which are not systematic surveys but rely on individuals to recognize a basking shark and know to report it. Research in 2018 in Inishowen demonstrated that a large majority of people (both tourists and residents) were unaware that they could report a sighting (Gray et al. 2022), making the chance of an unreported sighting throughout the model time period quite high. While the IBSG/IWDG data show that fewer sharks were reported pre-2000 than post-2000, this may not be related to abundance but instead may be attributed to the availability and widespread use of cell phones, especially smartphones, which rapidly increased in Ireland in the 2010s (Google 2013; McInturf et al. 2026). Publicly reported sightings also lack data on true absences. Future systematic surveys can help provide a more comparable dataset.

This model is also limited by its focus on surfacing sharks. This is because the sightings data face the same limitation. However, advances in technology, including increasingly complex and data-heavy tags, will result in a deeper understanding of basking sharks’ subsurface behavior. This IBM could then be expanded to incorporate subsurface behavior, subsurface preyscapes, and bathymetric data, allowing for more holistic ecological modeling. In the future, this model can be duplicated in other areas where basking sharks are sighted, with moderate changes to the underlying map and zooplankton data.

As data on shark populations improve, the model can be improved by estimating death and birth rates, the age or sex of the sharks, and the number of sharks that are likely to exhibit site fidelity in the area. While these data are largely unavailable at the moment, research is currently underway to expand the knowledge base of sharks in Ireland.

Conclusion

Understanding how food drives both long-term migratory patterns and localized shark behavior is vital for conservation in the face of climate change, and individual-based modeling is a mechanism by which theories of behavior can be tested and, with improved environmental data, predictive models can be generated. Long-term sea surface temperature (SST) increases have been observed in the North Atlantic in previous decades, and multiple studies have found that this increase in SST has led to a northward shift in zooplankton communities, with an increase in the range of warm-water zooplankton and a decrease in the range of cold-water zooplankton, including C. finmarchicus (Gregory et al. 2009; Heneghan et al. 2023). At a smaller scale, C. helgolandicus may be replacing C. finmarchicus as a result of climate change (Wilson et al. 2016). Climate change may also have complex impacts on copepod behavior, nutrient storage, the location of blooms, and surface patchiness, further impacting their predators (Folt and Burns 1999; Feng et al. 2014; Melle et al. 2014; Bresnan et al. 2015). Understanding how individual sharks make both localized and large-scale movement decisions will assist in the development of climate-resilient conservation strategies.

Even with limited environmental data, this IBM was able to reproduce long-term trends found in publicly reported sightings. This model allows for flexibility in both environmental modeling and behavioral attributes, allowing for increasing complexity as new data become available, as well as for the testing of different hypotheses of behavior. Environmental research is still required to bolster the underlying environment, thereby creating even more realistic behaviors. The ability of IBMs to account for individualized shark behavior may become more necessary in light of current research focused on individualized social behavior of basking sharks (A. McInturf, personal communication, 8 August 2023), as well as other research demonstrating unique social groupings in other sharks (Marosi et al. 2026). As the first IBM of basking sharks, this research demonstrated the possibility of modeling basking shark behavior via an IBM and provides avenues for future research.

Acknowledgments

The authors thank the Irish Basking Shark Group and Irish Whale and Dolphin Group for sharing the long-term sightings datasets. The authors also thank the Jackson Laboratory and George Mason University for computing resources.

References

  • Anderson SD, Chapple TK, Jorgensen SJ, Klimley AP, Block BA (2011) Long-term individual identification and site fidelity of white sharks, Carcharodon carcharias, off California using dorsal fins. Marine Biology 158(6): 1233–1237. https://doi.org/10.1007/s00227-011-1643-5
  • Austin RA, Hawkes LA, Doherty PD, Henderson SM, Inger R, Johnson L, Pikesley SK, Solandt J-L, Speedie C, Witt MJ (2019) Predicting habitat suitability for basking sharks (Cetorhinus maximus) in UK waters using ensemble ecological niche modelling. Journal of Sea Research 153: 101767. https://doi.org/10.1016/j.seares.2019.101767
  • Baduini C (1995) Feeding ecology of the basking shark (Cetorhinus maximus) relative to distribution and abundances of prey. Master’s Theses, 972 pp. https://doi.org/10.31979/etd.zh8m-jsdr
  • Berrow S, Heardman C (1994) The Basking Shark Cetorhinus maximus (Gunnerus) in Irish Waters: Patterns of Distribution and Abundance. Biology and Environment: Proceedings of the Royal Irish Academy 94B: 101–107.
  • Berrow S, De Boeck M, Davies RA, Ryan F, Collins R (2025) Evidence for successive breaching events by basking sharks Cetorhinus maximus (Lamniformes: Cetorhinidae). Environmental Biology of Fishes. https://doi.org/10.1007/s10641-025-01706-7
  • Braun CD, Skomal GB, Thorrold SR (2018) Integrating Archival Tag Data and a High-Resolution Oceanographic Model to Estimate Basking Shark (Cetorhinus maximus) Movements in the Western Atlantic. Frontiers in Marine Science 5(25). https://doi.org/10.3389/fmars.2018.00025
  • Bresnan E, Cook KB, Hughes SL, Hay SJ, Smith K, Walsham P, Webster L (2015) Seasonality of the plankton community at an east and west coast monitoring site in Scottish waters. Journal of Sea Research 105: 16–29. https://doi.org/10.1016/j.seares.2015.06.009
  • Chapple TK, Cade DE, Goldbogen J, Massett N, Payne N, McInturf AG (2024) Behavioral response of megafauna to boat collision measured via animal-borne camera and IMU. Frontiers in Marine Science 11. https://doi.org/10.3389/fmars.2024.1430961
  • Chapuis L, Collin SP, Yopak KE, McCauley RD, Kempster RM, Ryan LA, Schmidt C, Kerr CC, Gennari E, Egeberg CA, Hart NS (2019) The effect of underwater sounds on shark behaviour. Scientific Reports 9(1): 6924. https://doi.org/10.1038/s41598-019-43078-w
  • Collin SP (2012) The Neuroecology of Cartilaginous Fishes: Sensory Strategies for Survival. Brain, Behavior and Evolution 80(2): 80–96. https://doi.org/10.1159/000339870
  • Cotton P, Sims DW, Fanshawe S, Chadwick M (2005) The effects of climate variability on zooplankton and basking shark (Cetorhinus maximus) relative abundance off southwest Britain. Fisheries Oceanography 14(2): 151–155. https://doi.org/10.1111/j.1365-2419.2005.00331.x
  • Crowe LM, O’Brien O, Curtis TH, Leiter SM, Kenney RD, Duley P, Kraus SD (2018) Characterization of large basking shark Cetorhinus maximus aggregations in the western North Atlantic Ocean. Journal of Fish Biology 92(5): 1371–1384. https://doi.org/10.1111/jfb.13592
  • Curtis TH, Robinson J, Pratt HL, Skomal GB, Whitney NM (2024) Novel pre‐copulatory behavior in basking sharks observed by drone. Journal of Fish Biology 105(4): 1333–1337. https://doi.org/10.1111/jfb.15858
  • Doherty PD, Baxter JM, Gell FR, Godley BJ, Graham RT, Hall G, Hall J, Hawkes LA, Henderson SM, Johnson L, Speedie CD, Witt MJ (2017a) Long-term satellite tracking reveals variable seasonal migration strategies of basking sharks in the north-east Atlantic. Scientific Reports (Nature Publisher Group); London 7: 42837. https://doi.org/10.1038/srep42837
  • Doherty PD, Baxter JM, Godley BJ, Graham RT, Hall G, Hall J, Hawkes LA, Henderson SM, Johnson L, Speedie C, Witt MJ (2017b) Testing the boundaries: Seasonal residency and inter-annual site fidelity of basking sharks in a proposed Marine Protected Area. Biological Conservation 209: 68–75. https://doi.org/10.1016/j.biocon.2017.01.018
  • Druon J-N, Hélaouët P, Beaugrand G, Fromentin J-M, Palialexis A, Hoepffner N (2019) Satellite-based indicator of zooplankton distribution for global monitoring. Scientific Reports 9(1): 4732. https://doi.org/10.1038/s41598-019-41212-2
  • Dyomin V, Davydova A, Morgalev S, Kirillov N, Olshukov A, Polovtsev I, Davydov S (2020) Monitoring of Plankton Spatial and Temporal Characteristics With the Use of a Submersible Digital Holographic Camera. Frontiers in Marine Science 7: 653. https://doi.org/10.3389/fmars.2020.00653
  • Elcock JN, Arostegui MC, McDonnell LH, Klöcker CA, Skomal GB, Thorrold SR, Braun CD (2026) Basking sharks overlap with primary and secondary deep scattering layers during overwintering migration in the Northwest Atlantic Ocean. PLoS ONE 21(6): e0348589. https://doi.org/10.1371/journal.pone.0348589
  • Feng J, Stige L, Durant J, Hessen D, Zhu L, Hjermann D, Llope M, Stenseth N (2014) Large-scale season-dependent effects of temperature and zooplankton on phytoplankton in the North Atlantic. Marine Ecology Progress Series 502: 25–37. https://doi.org/10.3354/meps10724
  • Gore M, Abels L, Wasik S, Saddler L, Ormond R (2018) Are close-following and breaching behaviours by basking sharks at aggregation sites related to courtship? Journal of the Marine Biological Association of the United Kingdom 99(3): 681–693. https://doi.org/10.1017/S0025315418000383
  • Gray C, Peters-Burton E, Smith C, Parsons ECM (2022) Basking shark tourism in Donegal, Ireland – a case-study of public interest and support for shark conservation. Aquatic Conservation: Marine and Freshwater Ecosystems 32(3): 537–550. https://doi.org/10.1002/aqc.3774
  • Grimm V, Berger U, Bastiansen F, Eliassen S, Ginot V, Giske J, Goss-Custard J, Grand T, Heinz SK, Huse G, Huth A, Jepsen JU, Jørgensen C, Mooij WM, Müller B, Pe’er G, Piou C, Railsback SF, Robbins AM, Rossmanith E, Rüger N, Strand E, Souissi S, Stillman RA, Vabø R, Visser U, DeAngelis DL (2006) A standard protocol for describing individual-based and agent-based models. Ecological Modelling 198(1): 115–126. https://doi.org/10.1016/j.ecolmodel.2006.04.023
  • Grimm V, Augusiak J, Focks A, Frank BM, Gabsi F, Johnston AS, Liu C, Martin BT, Meli M, Radchuk V, Thorbek P, Railsback SF (2014) Towards better modelling and decision support: Documenting model development, testing, and analysis using TRACE. Ecological Modelling 280: 129–139. https://doi.org/10.1016/j.ecolmodel.2014.01.018
  • Hazen EL, Scales KL, Maxwell SM, Briscoe DK, Welch H, Bograd SJ, Bailey H, Benson SR, Eguchi T, Dewar H, Kohin S, Costa DP, Crowder LB, Lewison RL (2018) A dynamic ocean management tool to reduce bycatch and support sustainable fisheries. Science Advances 4(5): eaar3001. https://doi.org/10.1126/sciadv.aar3001
  • Heneghan RF, Everett JD, Blanchard JL, Sykes P, Richardson AJ (2023) Climate-driven zooplankton shifts cause large-scale declines in food quality for fish. Nature Climate Change 13(5): 470–477. https://doi.org/10.1038/s41558-023-01630-7
  • Hoelzel R, Shivji M, Magnussen J, Francis M (2006) Low worldwide genetic diversity in the basking shark (Cetorhinus maximus). Biology Letters 2(4): 639–642. https://doi.org/10.1098/rsbl.2006.0513
  • Hoogenboom JL, Wong SNP, Ronconi RA, Koopman HN, Murison LD, Westgate AJ (2015) Environmental predictors and temporal patterns of basking shark (Cetorhinus maximus) occurrence in the lower Bay of Fundy, Canada. Journal of Experimental Marine Biology and Ecology 465: 24–32. https://doi.org/10.1016/j.jembe.2015.01.005
  • Jabado RW, García-Rodríguez E, Kyne PM, Palacios MDM, Charles R, Bettcher Brito V, Rohner CA, O’Connor B, Armstrong AO, Batlle-Morera A, Gonzalez-Pestana A (2025) European Atlantic: A regional compendium of Important Shark and Ray Areas. IUCN SSC Shark Specialist Group. https://doi.org/10.59216/ssg.isra.2025.r2
  • Johnston EM, Halsey LG, Payne NL, Kock AA, Iosilevskii G, Whelan B, Houghton JDR (2018) Latent power of basking sharks revealed by exceptional breaching events. Biology Letters 14(9): 20180537. https://doi.org/10.1098/rsbl.2018.0537
  • Johnston EM, Mayo PA, Mensink PJ, Savetsky E, Houghton JDR (2019) Serendipitous re-sighting of a basking shark Cetorhinus maximus reveals inter-annual connectivity between American and European coastal hotspots. Journal of Fish Biology 95(6): 1530–1534. https://doi.org/10.1111/jfb.14163
  • Klöcker CA, Vihtakari M, Arostegui MC, Schlindwein A, Ferter K, Bjelland O, Dolton HR, Langangen Ø, Queiroz N, Sims DW, Junge C (2026) Wide-ranging, year-round breaching behaviour of basking sharks revealed by long-term biologging. Biology Letters 22(4): 20250696. https://doi.org/10.1098/rsbl.2025.0696
  • Lieber L, Hall G, Hall J, Berrow S, Johnston E, Gubili C, Sarginson J, Francis M, Duffy C, Wintner SP, Doherty PD, Godley BJ, Hawkes LA, Witt MJ, Henderson SM, de Sabata E, Shivji MS, Dawson DA, Sims DW, Jones CS, Noble LR (2020) Spatio-temporal genetic tagging of a cosmopolitan planktivorous shark provides insight to gene flow, temporal variation and site-specific re-encounters. Scientific Reports 10(1): Article 1. https://doi.org/10.1038/s41598-020-58086-4
  • Marosi ND, Ellis S, Jacoby DMP, Brunnschweiler JM, Croft DP (2026) Rolling in the deep: Drivers of social preferences and social interactions within a bull shark aggregation in Fiji. Animal Behaviour 234: 123511. https://doi.org/10.1016/j.anbehav.2026.123511
  • McInturf AG, Muhling B, Bizzarro JJ, Fangue NA, Ebert DA, Caillaud D, Dewar H (2022) Spatial Distribution, Temporal Changes, and Knowledge Gaps in Basking Shark (Cetorhinus maximus) Sightings in the California Current Ecosystem. Frontiers in Marine Science 9. https://www.frontiersin.org/articles/10.3389/fmars.2022.818670
  • McInturf AG, Bowman J, Schulte JM, Newton KC, Vigil B, Honig M, Pelletier S, Cox N, Lester O, Cantor M, Chapple TK (2023) A unified paradigm for defining elasmobranch aggregations. ICES Journal of Marine Science, fsad099. https://doi.org/10.1093/icesjms/fsad099
  • McInturf AG, Gray C, Phillips N, Leeney R, Griffin D, Lynam C, Bortoluzzi J, Loose ED, Vinu-Srivatsan S, Massett N, Berrow S, Whooley P, Johnston E (2026) Basking shark (Cetorhinus maximus) sightings data over fifty years informs species distribution models, temporal trends, and space use patterns in Ireland. Endangered Species Research 60: esr01515. https://doi.org/10.3354/esr01515
  • Melle W, Runge J, Head E, Plourde S, Castellani C, Licandro P, Pierson J, Jonasdottir S, Johnson C, Broms C, Debes H, Falkenhaug T, Gaard E, Gislason A, Heath M, Niehoff B, Nielsen TG, Pepin P, Stenevik EK, Chust G (2014) The North Atlantic Ocean as habitat for Calanus finmarchicus: Environmental factors and life history traits. Progress in Oceanography 129: 244–284. https://doi.org/10.1016/j.pocean.2014.04.026
  • Montero-Quintana AN, Ocampo-Valdez CF, Vázquez-Haikin JA, Sosa-Nishizaki O, Osorio-Beristain M (2021) Whale shark (Rhincodon typus) predatory flexible feeding behaviors on schooling fish. Journal of Ethology 39(3): 399–410. https://doi.org/10.1007/s10164-021-00717-y
  • Murphy KJ, Ciuti S, Kane A (2020) An introduction to agent-based models as an accessible surrogate to field-based research and teaching. Ecology and Evolution 10(22): 12482–12498. https://doi.org/10.1002/ece3.6848
  • National Research Council (US) Committee on Potential Impacts of Ambient Noise in the Ocean on Marine Mammals (2003) Effects of Noise on Marine Mammals. In Ocean Noise and Marine Mammals. National Academies Press. https://www.ncbi.nlm.nih.gov/books/NBK221255/
  • Nevitt GA, Bonadonna F (2005) Sensitivity to dimethyl sulphide suggests a mechanism for olfactory navigation by seabirds. Biology Letters 1(3): 303–305. https://doi.org/10.1098/rsbl.2005.0350
  • Okey T, Wallace S, Gallucci V (2010) Shark aggregation in coastal waters of British Columbia. Marine Ecology Progress Series 414: 249–256. https://doi.org/10.3354/meps08718
  • OSPAR Commission (2015) Background Document for Basking Shark Cetorhinus maximus. Convention for the Protection of the Marine Environment of the North-East Atlantic.
  • Owen K, Saeki K, Warren JD, Bocconcelli A, Wiley DN, Ohira S-I, Bombosch A, Toda K, Zitterbart DP (2021) Natural dimethyl sulfide gradients would lead marine predators to higher prey biomass. Communications Biology 4(1): 149. https://doi.org/10.1038/s42003-021-01668-3
  • Papastamatiou YP, Bodey TW, Caselle JE, Bradley D, Freeman R, Friedlander AM, Jacoby DMP (2020) Multiyear social stability and social information use in reef sharks with diel fission–fusion dynamics. Proceedings of the Royal Society B: Biological Sciences 287(1932): 20201063. https://doi.org/10.1098/rspb.2020.1063
  • Pilkey OH [with Pilkey-Jarvis L] (2007) Useless arithmetic: Why environmental scientists can’t predict the future. Columbia University Press.
  • Planque B, Aarflot JM, Buttay L, Carroll J, Fransner F, Hansen C, Husson B, Langangen Ø, Lindstrøm U, Pedersen T, Primicerio R, Sivel E, Skogen MD, Strombom E, Stige LC, Varpe Ø, Yoccoz NG (2022) A standard protocol for describing the evaluation of ecological models. Ecological Modelling 471: 110059. https://doi.org/10.1016/j.ecolmodel.2022.110059
  • Rigby CL, Barreto R, Carlson J, Fernando D, Fordham S, Francis MP, Herman K, Jabado RW, Liu KM, Marshall A, Romanov E, Kyne PM (2021) Cetorhinus maximus (amended version of 2019 assessment) [Dataset]. In The IUCN Red List of Threatened Species 2021: e.T4292A194720078). IUCN. https://doi.org/10.2305/IUCN.UK.2021-1.RLTS.T4292A194720078.en
  • Rudd JL, Bartolomeu T, Dolton HR, Exeter OM, Kerry C, Hawkes LA, Henderson SM, Shirley M, Witt MJ (2021a) Basking shark sub-surface behaviour revealed by animal-towed cameras. PLoS ONE 16(7): e0253388. https://doi.org/10.1371/journal.pone.0253388
  • Rudd JL, Exeter OM, Hall J, Hall G, Henderson SM, Kerry C, Witt MJ, Hawkes LA (2021b) High resolution biologging of breaching by the world’s second largest shark species. Scientific Reports 11(1): Article 1. https://doi.org/10.1038/s41598-021-84670-3
  • Schmolke A, Thorbek P, DeAngelis DL, Grimm V (2010) Ecological models supporting environmental decision making: A strategy for the future. Trends in Ecology & Evolution 25(8): 479–486. https://doi.org/10.1016/j.tree.2010.05.001
  • Shunmugapandi R, McCarry CL, McKee D, Mitchell C (2025) Ocean color anomaly detection to estimate surface Calanus finmarchicus concentration in the Gulf of Maine. Frontiers in Marine Science 12: 1507638. https://doi.org/10.3389/fmars.2025.1507638
  • Sims DW (1999) Threshold foraging behaviour of basking sharks on zooplankton: Life on an energetic knife-edge? Proceedings of the Royal Society B: Biological Sciences 266(1427): 1437–1443. https://doi.org/10.1098/rspb.1999.0798
  • Sims DW, Quayle VA (1998) Selective foraging behaviour of basking sharks on zooplankton in a small-scale front. Nature 393(6684): 460–464. https://doi.org/10.1038/30959
  • Sims DW, Reid PC (2002a) Congruent trends in long‐term zooplankton decline in the north‐east Atlantic and basking shark (Cetorhinus maximus) fishery catches off west Ireland. Fisheries Oceanography 11(1): 59–63. https://doi.org/10.1046/j.1365-2419.2002.00189.x
  • Sims DW, Reid PC (2002b) Congruent trends in long-term zooplankton decline in the north-east Atlantic and basking shark (Cetorhinus maximus) fishery catches off west Ireland. Fisheries Oceanography 11(1): 59–63. https://doi.org/10.1046/j.1365-2419.2002.00189.x
  • Sims DW, Southall EJ, Quayle VA, Fox AM (2000) Annual social behaviour of basking sharks associated with coastal front areas. Proceedings of the Royal Society B: Biological Sciences 267(1455): 1897–1904. https://doi.org/10.1098/rspb.2000.1227
  • Sims DW, Southall EJ, Merrett DA, Sanders J (2003) Effects of zooplankton density and diel period on surface-swimming duration of basking sharks. Journal of the Marine Biological Association of the UK 83(3): 643–646. https://doi.org/10.1017/S0025315403007598h
  • Sims DW, Berrow S, O’Sullivan KM, Pfeiffer NJ, Collins R, Smith KL, Pfeiffer BM, Connery P, Wasik S, Flounders L, Queiroz N, Humphries NE, Womersley FC, Southall EJ (2022) Circles in the sea: Annual courtship “torus” behaviour of basking sharks Cetorhinus maximus identified in the eastern North Atlantic Ocean. Journal of Fish Biology 101(5): 1–22. https://doi.org/10.1111/jfb.15187
  • Skomal G, Zeeman S, Chisholm J, Summers E, Walsh H, McMahon K, Thorrold S (2009) Transequatorial Migrations by Basking Sharks in the Western Atlantic Ocean. Current Biology 19(12): 1019–1022. https://doi.org/10.1016/j.cub.2009.04.019
  • Speedie CD, Johnson LA, Witt MJ (2009) Basking Shark Hotspots on the West Coast of Scotland: Key sites, threats and implications for conservation of the species (Commissioned Report No.339).
  • Squazzoni F, Boero R (2010) Complexity-Friendly Policy Modelling. In Innovation in Complex Social Systems, edited by Petra Ahrweiler. New York: Routledge, 290–299.
  • Thorburn J, Collins PC, Garbett A, Vance H, Phillips N, Drumm A, Cooney J, Waters C, Ó’Maoiléidigh N, Johnston E, Dolton HR, Berrow S, Hall G, Hall J, Delvillar D, McGill R, Whoriskey F, Fangue NA, McInturf AG, Rypel AL, Kennedy R, Lilly J,. Rodger JR, Adams CE, van Geel NCF, Risch D, Wilkie L, Henderson S, Mayo PA, Mensink PJ, Witt MJ, Hawkes LA, Klimley AP, Houghton JDR (2024) Assessing the potential of acoustic telemetry to underpin the regional management of basking sharks (Cetorhinus maximus). Animal Biotelemetry 12(1): 20. https://doi.org/10.1186/s40317-024-00370-5
  • Wilensky U (1999) NetLogo. Center for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL. https://www.netlogo.org
  • Wilson RJ, Heath MR, Speirs DC (2016) Spatial Modeling of Calanus finmarchicus and Calanus helgolandicus: Parameter Differences Explain Differences in Biogeography. Frontiers in Marine Science 3. https://doi.org/10.3389/fmars.2016.00157

Additional information

Conflict of interest

The authors have declared that no competing interests exist.

Ethical statement

No ethical statement was reported.

Artificial Intelligence (AI) use

The authors accept full responsibility for the content of the manuscript, including the disclosure of any use of AI.

No AI tools were used in the preparation of this manuscript.

Funding

No funding was reported.

Author contributions

Conceptualization: CS, EPB, CG, EP. Data curation: ACG, CG. Formal analysis: CG. Investigation: DSR. Methodology: CG, ACG, DSR. Software: ACG, DSR. Supervision: EP, DSR, EPB, CS. Validation: CG. Visualization: CG. Writing - original draft: CG. Writing - review and editing: CG, CS, ACG, DSR, EP, EPB.

Author ORCIDs

C. Gray https://orcid.org/0000-0001-9441-7242

D.S. Rothman https://orcid.org/0000-0003-1479-2849

E. Peters-Burton https://orcid.org/0000-0003-1233-9119

E.C.M Parsons https://orcid.org/0000-0002-0464-1046

A. Garretson https://orcid.org/0000-0002-7260-0131

Data availability

All of the data that support the findings of this study are available in the main text or Supplementary Information.

1 Due to a lack of CPR data from 26 October–1 November 2018, the model ends the entire run on 26 October 2018 instead of 31 October 2018.
2 CPR only identifies Pseudocalanus to the genus level (Richardson et al. 2006).
3 It is likely a combination of smell and sound (Sims and Quayle 1998). The distance that smell or sound travels in water can vary depending on a variety of factors that the model does not currently account for (Collin 2012; Dove and Pierce 2021).
4 Sharks cannot utilize their memory of high-zooplankton patches until they have a list of six or more patches that had three times the threshold level of zooplankton.
5 Note that the model merely records whether the shark has eaten. The actual consumption of zooplankton is not represented in the model.
6 The shark is set to “hidden,” is no longer visible to other sharks, and does not participate in daily decision-making.
7 This was done because, during model development, it was observed that if sharks did not leave a patch because of insufficient zooplankton, they stopped moving for the duration of the model run once they found another shark.
8 The decision was made to have sharks search for food first, as this aligned better with current real-world research, which theorizes that zooplankton is a main driver of shark movements (Cotton et al. 2005; Sims et al. 2000, 2003; Sims and Reid 2002b). Model versions in which sharks first searched for sharks, then zooplankton, were tested, but there was no difference in the model results.
9 Across the entire 1982–2018 period, there were 34 sighting reports in the last week of March, the majority of which were between 27 and 31 March. For data analysis, these were counted in the first week of April. There were 14 other sighting reports made outside this time frame between 1982 and 2018, which were not used in the data analysis.
10 There was little difference among the RMSE, ME, and MAE results when comparing Inishowen with all of Ireland, so it was not necessary to give one greater weight than the other.

Supplementary materials

Supplementary material 1 

Basking Shark IBM

Chelsea Gray, Dale S. Rothman, Erin Peters-Burton, Cynthia Smith, E.C.M Parsons, Alexis Garretson

Data type: zip

Explanation note: Contains IBM (netlogo file) and necessary input data to run the model.

This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Download file (227.99 kb)
Supplementary material 2 

Model documentation

Chelsea Gray, Dale S. Rothman, Erin Peters-Burton, Cynthia Smith, E.C.M Parsons, Alexis Garretson

Data type: docx

This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Download file (458.86 kb)
Supplementary material 3 

Supplemental model results

Chelsea Gray, Dale S. Rothman, Erin Peters-Burton, Cynthia Smith, E.C.M Parsons, Alexis Garretson

Data type: docx

This dataset is made available under the Open Database License (http://opendatacommons.org/licenses/odbl/1.0/). The Open Database License (ODbL) is a license agreement intended to allow users to freely share, modify, and use this Dataset while maintaining this same freedom for others, provided that the original source and author(s) are credited.
Download file (41.43 kb)
login to comment